Event processing systems are coming into increasingly widespread use in a variety of different enterprises. However, existing event processing systems have a number of significant drawbacks. For example, such systems often require custom-built transformations for each different type of input event information to be processed, and are generally unable to detect, normalize and combine event information from multiple federated information sources. Also, the existing event processing systems are typically very limited in terms of the complexity of events that can be recognized and processed.
A given enterprise may therefore have to combine multiple disparate event processing systems in order to handle complex events. In addition, exception handling in such arrangements often requires extensive human intervention. As a result, the deployment of complex event processing functionality can be unduly expensive and time-consuming for the enterprise.